REJECT H0 The p-value 0.00 < 0.05. Thus, H0 can be rejected. Average customer satisfaction with HP is different between customers with high levels of tech support and those with low or medium levels of tech support satisfaction (same values ( Correlation analysis (Relational, metric, association between 2 variables ( · Measures the degree to which there is a linear association between two interval or ratio scaled variables RQ: Is there a correlation between pricing satisfaction and overall customer satisfaction ? H0: p = 0 There is no linear association between satisfaction with pricing and perceived quality of service H1: p ≠ 0 There is an association between satisfaction with contract requirement and perceived quality of service Interpretation
FAIL TO REJECT H0 The p-value 0.60 > 0.05. Thus, H0 cannot be rejected. There is no linear association between overall customer dissatisfaction and satisfaction with pricing for HP customers REJECT H0 The p-value 0.00 < 0.05. Thus, H0 can be rejected. There is a linear association between overall customer satisfaction and pricing satisfaction. Higher pricing satisfaction correlates with lower overall customer dissatisfaction (or higher customer satisfaction). The association is negative and strong (-0.454 ( ** Strong correlation does not mean IV changes DV by a lot, it is not how much DV changes but how precisely can IV predict DV (high correlation/association ( Regression (connecting both relational and comparative ( · Explain the variation in dependent variables (outcome variables) using other metric variables as independent variables (predictors), and/or test for differences across groups (using dummy variables) · CAN BE MULTIPLE OR JUST ONE OK RQ: Is there an association between _______ (IV) and __________ (DV ?(
H0:B=0 There is no association between _________ (IV) and __________ (DV ( H1:B ≠ 0 There is an association between __________ (IV) and __________ (DV ( Additional hypothesis to test significance of overall regression : H0:B1=B2=B3=0 The model is invalid H1: at least one of the B1, B2, and B3 is not 0
EXAMPLE FOR REGRESSION
Interpret: p-value, R2 (validity), p-value again, one unit increase in ____ corresponds with _____ increase/decrease in perceived quality, directionally which is better predictor? (SC beta ( MODEL VALIDITY · The p-value 0.00 < 0.05. Thus, H0 can be rejected. The model is valid. · The R square is 0.268. This means we can interpret 26.8% of the variability in _____ (DV) ANALYSIS The IV ____ and ____ has an effect on _____ (DV (
FAIL TO REJECT H0 Do not reject ? 6 since p-value 0.60 > 0.05. Satisfaction with calling plans has no effect on perceived quality of service REJECT H0 Reject ? 7 since p-value 0.000 < 0.05. One unit increase in satisfaction with billing statement corresponds with 0.278 in perceived quality of service . · Directionally, satisfaction with contract requirement is a better predictor of DV than satisfaction with billing statement (0.341 vs 0.258) -> SC Beta Segmentation · Identify key market segments
* A priori is YOU KNOW, post hoc is when YOU DON’T KNOW so you find the segments using segmentation analysis Regression by segment / a priori Example RQ: Is there a difference in average overall customer satisfaction with HP between
You've reached the end of your free preview.
Want to read all 44 pages?
- Two '16